#A long-term restoration experiment at Rancho Marino Reserve.
Purpose
To explore existing evidence and data empirically and assess future research activities including sampling for efficicacy of intervention on recruitment of natives into coastal grassland with extensive exotic plant representation.
Methods
The control macroplots were not used.
From data sheet columns: Plot 12x12m(12 setup & originally sampled, but as you noted control sites weren’t uses) Subplot 3x3m (4 per plot)
Quadrat 1x1m (5 per Subplot)
Macro meta-data
Don Canestro In summary, 4 years of sampling 1 before and three after planning. I also have data on plug grow out and which spp. seeds had best germination rates and plot images from 2006 and 2009.
Between 2002 before data and 2006 data collection was kill time. I was told by restoration experts 3 years of death on a plot was good to remove seed bank and reduce competition
Below are descriptions of the attached files. “harding grassexpt3_05.doc” describes the experiment “HARDINGGRASSPLOTS.xls” illustrates the plot layout. “Percent cover data2002.xls” is before treatment data Other files self evident.
Data & basic library load
library(tidyverse)
library(plotly)
#2009 only
data <- read_csv("data/2009.data.csv")
data## # A tibble: 160 × 73
## data plot plot.treatment subplot subplot.treatment.number
## <int> <int> <chr> <chr> <int>
## 1 2009 1 plastic D 1
## 2 2009 1 plastic D 1
## 3 2009 1 plastic D 1
## 4 2009 1 plastic D 1
## 5 2009 1 plastic D 1
## 6 2009 2 till.spray A 1
## 7 2009 2 till.spray A 1
## 8 2009 2 till.spray A 1
## 9 2009 2 till.spray A 1
## 10 2009 2 till.spray A 1
## # ... with 150 more rows, and 68 more variables: subplot.treatment <chr>,
## # quadrat <int>, max.height <int>, tallest.species <chr>, `small mammal
## # holes` <int>, bare <int>, thatch <int>, Aica <int>, Anar <int>,
## # Avena <int>, Baca <int>, Bado <int>, Bapi <int>, Brca <int>,
## # Brdi <int>, Brho <int>, Brmi <int>, Care <int>, Carp <int>,
## # Cheno <int>, Civu <int>, Coca <int>, Con <int>, Cyper <int>,
## # Cyper0sp2 <int>, Daca <int>, Elgl <int>, Eltr <int>, flat0sedge <int>,
## # Gedi <int>, Ger0sp <int>, Gna <int>, Hepu <int>, Hobr <int>,
## # Hypo <int>, Isce <int>, Jubu <int>, Jupa <int>, Junc.sp <int>,
## # Junc.sp.2 <int>, Juoc <int>, Lomu <int>, Lope <int>, Lyhy <int>,
## # Malva <int>, Mepo <int>, Myopo <int>, Napu <int>, Phaq <int>,
## # Poly <int>, Rasa <int>, Ruac <int>, Rucr <int>, Sil <int>, Sima <int>,
## # Sonch <int>, Spma <int>, Trifol0sp <int>, forb <int>, grass <int>,
## # Vela <int>, Ver <int>, Vumy <int>, total.cover <int>,
## # density.Daca <int>, density.Elgl <int>, density.Napu <int>,
## # total.native.density <int>#str(data)EDA & Data Viz
#totals
p <- ggplot (data, aes(plot.treatment, total.cover)) + geom_boxplot() + facet_wrap(~subplot.treatment)
ggplotly(p)p <- ggplot (data, aes(plot.treatment, total.native.density)) + geom_boxplot() + facet_wrap(~subplot.treatment)
ggplotly(p)p <- ggplot (data, aes(plot.treatment, total.native.density, color = subplot.treatment)) + geom_boxplot() + geom_point()
ggplotly(p)#for each native
p <- ggplot (data, aes(plot.treatment, density.Daca)) + geom_boxplot() + facet_wrap(~subplot.treatment)
ggplotly(p)p <- ggplot (data, aes(plot.treatment, density.Elgl)) + geom_boxplot() + facet_wrap(~subplot.treatment)
ggplotly(p)p <- ggplot (data, aes(plot.treatment, density.Napu)) + geom_boxplot() + facet_wrap(~subplot.treatment)
ggplotly(p)#nested GLM simple
m <- glm(total.native.density ~plot.treatment/subplot.treatment, family = poisson, data = data)
summary(m)##
## Call:
## glm(formula = total.native.density ~ plot.treatment/subplot.treatment,
## family = poisson, data = data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.8460 -2.3022 -0.0002 0.2717 8.9752
##
## Coefficients:
## Estimate Std. Error
## (Intercept) -1.730e+01 7.754e+02
## plot.treatmenttill.spray 3.301e-08 1.097e+03
## plot.treatmentplastic:subplot.treatmentculm 1.870e+01 7.754e+02
## plot.treatmenttill.spray:subplot.treatmentculm 1.804e+01 7.754e+02
## plot.treatmentplastic:subplot.treatmentplug 1.828e+01 7.754e+02
## plot.treatmenttill.spray:subplot.treatmentplug 1.842e+01 7.754e+02
## z value Pr(>|z|)
## (Intercept) -0.022 0.982
## plot.treatmenttill.spray 0.000 1.000
## plot.treatmentplastic:subplot.treatmentculm 0.024 0.981
## plot.treatmenttill.spray:subplot.treatmentculm 0.023 0.981
## plot.treatmentplastic:subplot.treatmentplug 0.024 0.981
## plot.treatmenttill.spray:subplot.treatmentplug 0.024 0.981
##
## (Dispersion parameter for poisson family taken to be 1)
##
## Null deviance: 903.42 on 159 degrees of freedom
## Residual deviance: 687.31 on 154 degrees of freedom
## AIC: 892.88
##
## Number of Fisher Scoring iterations: 15anova(m, test = "Chisq")## Analysis of Deviance Table
##
## Model: poisson, link: log
##
## Response: total.native.density
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 159 903.42
## plot.treatment 1 1.508 158 901.92 0.2194
## plot.treatment:subplot.treatment 4 214.608 154 687.31 <2e-16
##
## NULL
## plot.treatment
## plot.treatment:subplot.treatment ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#can also use lme4 package and nest with the + (1|factor) code